Frequently asked questions about cd2browse:
Question:
27.11.2009 Why does the Windows installer of cd2browse save lots of files to directories I have’nt been even asked for ?
Answer:
Because cd2browse is running on several platforms, files are stored in a unix like directory structure. Actually you can use the Unix tar ball and install the contained structure where ever you want, or you may move the directories after the installation of the win packages, but then the initial variable section of the two perl scripts needs to be modified. To make the installation on windows easier, the project uses CreateInstallFree from http://www.createinstall.com/ But this software has limited possibilities and therefore and because of the easier installation I chose fixed directories.
Question:
After the installation I found lots of files under /var/cd2browse – what are they used for ?
Answer:
This is the template section. As described in the manual , here you can modify the look and feel of the later cd2browse results. The provided templates are just a simple example.
Question:
26.11.2009 After installing Version 1.03 on Windows XP – lots of Problems:
A. Instead of asking about where to install (most installers have no difficulty in finding my directory C:Program Files), it unilaterally chose to make a new directory (C:programs), where it installed C:\programs\util\xpdf and C:programs\util\catdoc
Answer:
cd2browse is written in perl and needs several external modules and programs. Most of these programs derive from the unix world and they have no windows installer. So I had to decide whether to save these programs to fixed directories, or leave them out of the package. The choice to use the windows installer is up to the user, because there are three packages of cd2browse – all of them can be used. Using the windows installer including binaries makes the installation much easier, but with the described restrictions.
B. The batch file hit a lot of problems, all of which wizzed past on the screen. It exitted without leaving a log file!
Answer: I will take this as a good idea for the next release, to write a log file – thanks !
C. It autostarted without being asked, and came up with a perlTk screen (quite nice). But this did nothing. On closure, it left a dos window with the following message Name “main::HKEY_LOCAL_MACHINE” used only once: possible typo at C:Perlcd2browseCD2browse.pl line 276. Cant locate Module/Load/Conditional.pm in @INC (@INC contains: C:/Perl/lib C:/Perl/site/lib .) at index.pl line 28. BEGIN failed–compilation aborted at index.pl line 28.
Answer:
Autostart is not a implemented function, so you must have been started manually.The message “main::HKEY_LOCAL_MACHINE” is just a warning – this is normal and does not disturb the function – however I will prevent it from coming up in the next release. The last message is a error message, that the module “Load/Conditional” is not installed with your Perl distribution. The recommended ActivePerl version contains it. If you use a different version you can load this module from the internet using ppm (Perl packet manager – with ActivePerl) or “cpan” with other Perl distributions.
D. Running perl C:Perlcd2browseindex.pl gives the same message Cant locate Module/Load/Conditional.pm in @INC
Answer:
The same as above
Question:
I installed the full package 1.02 containing Perl, but after installing Perl and Image Magick, when I tried to run CD2browse.pl, I got a error message in the black console window like “cannot find Perlmagick” and the processing of the desired file tree did’nt work.
Answer:
There are two probable reasons for that:
1. You forgot to install Image magick with the activated option “Install PerlMagick for ActiveState Perl v5.8.8”
2. On some systems the installation of ImageMagick fails for the first time
In both cases, try to restart the installation of ImageMagick. The installation package is located under c:\Perl\cd2browse\software on your hard drive. Normally this will cure the problem.
Question:
I start index.pl, but after a short while I get the message: “unable to open file .html/000<number>.html – no such file or directory …
Answer:
Obviously the path provided with the -p option does not exist, or you just forgot to put the path into double quotes – this is necessary, when the path has spaces.
Question:
If I have different sizes of images to index, the distance in between index lines is very different, because of different thumbnail heights – this looks bad – is there a possibility to improve this.
Answer:
By default the te thumbnail size is 90×90 pixels. If you change for example to 120 x 90, the height is approx, the same for all images if the aspect ratio is 1:1,33 like on normal images. This means a mixture of landscape and portrait images will result in the same line distance.
Question:
I indexed a directory with approx. 30000 files and it worked so far, but now the search java script engine is very slow – means it seems that nothing happens and after a while I get a error message, that java script does not respond.
Answer:
Depending on the system speed, you may continue processing the calculation, and after a while it will show the search results. Because the search engine is based on java script, and the whole index file is loaded into memory, the speed is rather slow. This is due to limitations of JS. A reasonable search speed is achieved up to approx. 10000 indexed files. (2-5 seconds on a 2GHZ single processor system) Also the search speed depends on the file size of the original. Although the search engine uses a pre-calculated and compressed index, the number of results affects the search time.
Question:
When I point my browser onto index.html, I get a message, that some content has been blocked.
Answer:
Because cd2browse uses javascript to calculate the tree view, you need to allow java script in your browser. Therefore check your browser settings and set the security section accordingly.
Question:
After changing the template files to change the design of the CD, and after re-calculation of the menus, the design looks the same again.
Answer:
cd2browse checks if the main files exist before mirroring from the resoure source. If yes, the existing files in the target directory remain. If you want to change the design, you may change it in the target directory directly, or you delete the main files in the target directory before calculation of the index.
Question:
I got the following Error:
Cant stat /var/cd2browse/resources: No such file or directory
at c:/programme/perl/site/lib/File/Mirror.pm line 27
when starting index.pl with command:
from c:\Programme\perl\own >> perl index.pl -p c:Katalog
Answer:
cd2browse starts with copying some resources to the destination directory, such as the index.html, the results pages and necessary images, because the intention is to run it later from a CD or DVD. Therefore the directories .menu .resources .thumb and .html have to be filled with the master files.
With the above message, obviously the master directory (/var/cd2browse) does not exist. This directory is included in the zip file. If you did’nt extract it to the drive root it exists in a sub directory, and can therefore not be found. If you want the master directory somewhere else, you may change the configuration section in the beginning of the index.pl main script. The $datahome variable defines the resource path.
Question:
I have a bunch of PDF documents to index and the indexing starts as expected, but afte finishing the content of the PDF files is not searchable through the offline search engine – only a few words from the file name are in the search index. On an other installation I have seen the index is much bigger, even if I use the same pdf files.
Answer:
When xpdf is not installed or if it is not included in the path variable, it will not be started and therefore the content of the pdf file is not included in the search index. Make sure xpdf is running from the command line – you may test this by running xpdf directly from the command line according to the following sample command:
pdf2text /path_to_test_pdf/test.pdf
If nothing or error messages come back, the extraction of text out of pdf does not work.